E1163: Does Google care whether people actually click the links on your page?
I break down a $50,000 internal linking experiment across 15 websites, plus a Google patent that describes assigning different weights to links based on the probability that users will click them.
The findings challenge a lot of standard SEO thinking around internal links, PageRank dilution, link placement, and how authority flows through a site.
In this episode:
- A $50,000 internal linking experiment conducted across 15 websites
- How increasing internal links produced higher rankings and higher ad RPMs
- What happened when pages went from 3 - 4 internal links to 20+
- Why the tester found no ranking decline from adding more internal links
- How stronger internal linking eliminated "Crawled - currently not indexed" problems across his sites
- How he found five internal links pointing to a page was an important threshold for indexing
- Why internal links appearing earlier than other internal links on the page produced stronger ranking results
- The click-tracking data showing a clear correlation between internal link clicks and ranking improvements
- How internal linking based on topical similarity performed better than restricting links by categories or tags
- How opening internal links in the same tab versus a new tab made no ranking difference in his tests
- Dan Petrovic's PageRank Split Experiment involving a page with 4,225 followed outbound links
- How a test site jumped from position 79 to position 3 despite receiving a link from a page with thousands of outgoing links
- Google's patent "Ranking Documents Based on User Behavior and/or Feature Data"
- Google's patent language describing link weights based on the probability that a particular link will be selected
- How Google's patent describes collecting information about which links users selected
- How browsers or browser assistants are described as sources of user-behavior data
- Google's patent's example of recording actual clicks from one page to multiple linked pages
- How click instances are used to train a model predicting which links users are likely to select
- How links on the same page do not necessarily deserve the same weight
- How this connects with what we already know about Google using click data through NavBoost
- Why the most important internal links on your site should be links people actually want to click
The big takeaway: internal linking is not just about adding links and distributing PageRank.
Placement matters. Relevance matters. User behavior matters. And the links people click carry more weight.
⭐️ Black Hat World: I've Spent Over $50K on Internal Linking Experiments - https://www.blackhatworld.com/seo/ama-guide-ive-spent-over-50k-on-internal-linking-experiments-and-tests-this-year-here-is-everything-you-wanted-to-know-about-it.1628963/#post-18067348
⭐️ PageRank Split Experiment - https://dejanmarketing.com/pagerank-split-experiment/
⭐️ Google Patent: Ranking Documents Based on User Behavior and/or Feature Data - https://patents.google.com/patent/US9305099
💎 Compact Keywords - My SEO Course - Get paying customers through SEO - Clear step-by-step video breakdowns - SEO templates to be copied and adapted for your products and services: https://compactkeywords.com/
00:00 $50K Link Click Question
00:31 BlackHat World Tests
01:49 More Links More RPM
02:41 PageRank Dilution Myth
03:38 Indexing And Link Placement
04:17 Clicks And Similarity Rules
06:10 Google Patent On Clicks
07:21 Model Weights By Probability
08:57 What This Means For SEO
10:16 Wrap Up
The Edward Show. Your daily search engine optimization podcast: https://edwardsturm.com/the-edward-show/
#searchengineoptimization #seo #technicalseo #webdevelopment